MEPX
Chapter 7 of 10All chapters
  1. 1What correlation measures
  2. 2The five explanations
  3. 3Confounders
  4. 4Experiments
  5. 5Natural experiments
  6. 6Weighing causal evidence
  7. 7Selection effects
  8. 8Regression to the mean
  9. 9Language and reporting
  10. 10Acting under uncertainty

Chapter 7 of 10

Selection effects

Who ended up in the data.

How it distorts

If being in the sample depends on both variables, an association appears with no causal link. Studying only hospital patients, successful companies or people who replied all do this.

  • Survivorship is the most familiar form.
  • Online reviews come from the delighted and the furious.

Checking

Ask how each record came to be in the dataset. If the answer relates to the outcome you are studying, the analysis needs to account for it.